BACKGROUND: Although about 7 million people in the US population use marijuana at least weekly, there is a paucity of scientific data on persistent neurocognitive effects of marijuana use. OBJECTIVE: To determine if neurocognitive deficits persist in 28-day abstinent heavy marijuana users and if these deficits are dose-related to the number of marijuana joints smoked per week. METHODS: A battery of neurocognitive tests was given to 28-day abstinent heavy marijuana abusers. RESULTS: As joints smoked per week increased, performance decreased on tests measuring memory, executive functioning, psychomotor speed, and manual dexterity. When dividing the group into light, middle, and heavy user groups, the heavy group performed significantly below the light group on 5 of 35 measures and the size of the effect ranged from 3.00 to 4.20 SD units. Duration of use had little effect on neurocognitive performance. CONCLUSIONS: Very heavy use of marijuana is associated with persistent decrements in neurocognitive performance even after 28 days of abstinence. It is unclear if these decrements will resolve with continued abstinence or become progressively worse with continued heavy marijuana use. These patients were also found significant increases in resistance to flow in the anterior and middle cerebral arteries. It is not know at present if there is a direct correlation between the cognitive changes and the blood flow abnormailities.